CN111969916A - Multi-current-difference-updating model-free prediction current control method for permanent magnet synchronous motor - Google Patents

Abstract

The invention relates to a model-free predictive current control method for a multi-current-difference updated permanent magnet synchronous motor, which comprises the following steps of firstly, obtaining three-phase stator current, rotor electrical angle and actual rotating speed at the moment k through a current sensor and an encoder, and obtaining three-phase stator current at the moment k-1; then, the dq-axis stator currents at the k moment and the k-1 moment are obtained through coordinate transformation, and the dq-axis current difference at the k moment is obtained through a current difference calculation module; calculating current differences corresponding to the equidirectional and reverse voltage vectors according to the dq-axis current difference at the moment k and updating a current difference lookup table; then calculating a current predicted value at the k +1 moment through a model-free prediction equation; and finally, obtaining the optimal voltage vector of the driving inverter through the rolling optimization of the cost function. The invention can realize the simultaneous updating of a plurality of current differences in another control period, ensure the reliability of the current difference lookup table and ensure the reliability of the zero vector current difference by forced updating.

Description

Multi-current-difference-updating model-free prediction current control method for permanent magnet synchronous motor

Technical Field

The invention relates to a model-free predictive current control method for a multi-current-difference updated permanent magnet synchronous motor, and belongs to the field of motor driving and control.

Background

Permanent Magnet Synchronous Motors (PMSM) have the characteristics of small volume, high power and efficiency, low heat generation, high reliability and the like, and are widely applied to the fields of metallurgy, machining, servo control, automobiles and the like. The traditional control method of the PMSM has vector control and direct torque control, however, coordinate change of the vector control is complex, adjustment of PI parameters is complicated, low-speed performance of the direct torque control is poor, and real-time requirement is high, so that a Model Predictive Current Control (MPCC) which has a simple structure, can realize multi-target optimization and has a high response speed is widely concerned by researchers.

Generally, the MPCC predicts a current state at the next moment through a mathematical model of the motor, and then outputs an optimal switching state of the control system through rolling optimization of a cost function to guarantee the control performance of the system. However, the MPCC scheme requires a more accurate mathematical model to ensure the accuracy of the predicted current. Therefore, uncertainty of motor parameters during actual operation inevitably leads to degradation of system control performance. Therefore, a Model-free predictive current control (MFPCC) method is proposed.

The MFPCC algorithm replaces a current prediction model containing motor parameters and complex multiplication and division operations in the MPCC by the addition operation of the current difference and the current state stored in the current difference lookup table, so that the parameter robustness of the system can be greatly improved, and the calculation load of the system is reduced. However, the MFPCC algorithm requires a precise current difference to ensure the accuracy of the predicted current, and therefore, a higher current difference update frequency is required.

Disclosure of Invention

The technical problem is as follows: aiming at the prior art, the model-free prediction current control method for the permanent magnet synchronous motor with multiple current difference updates is provided, so that the update frequency of the current difference can be effectively improved, and the current difference corresponding to two different voltage vectors can be updated simultaneously in one control period.

The technical scheme is as follows: a permanent magnet synchronous motor model-free prediction current control method with multiple current difference updating comprises the following steps:

step 1: will give a rotation speed N_r ^refAnd the actual speed N measured by the encoder_rThe deviation between the two is obtained by a PI controller of a rotating speed loop to obtain a q-axis current reference value i at the moment of (k +1)_q ^ref(k +1) and giving a reference value i of the d-axis current_d ^ref(k+1)＝0；

Step 2: acquiring three-phase stator current i at (k-1) time and k time through a current sensor_s(k-1) and i_s(k) S ═ a, b, c; then obtaining stator currents of the (k-1) moment and the k moment alpha beta axis through Clark transformation; and then, acquiring the rotor electrical angle theta at the time k through an encoder, and calculating the dq-axis stator current i at the time (k-1) and the time k through Park change_d(k-1)、i_q(k-1) and i_d(k)、i_q(k)；

And step 3: will i_d(k-1)、i_q(k-1)、i_d(k) And i_q(k) Sending the current difference to a dq axis current difference calculation module to obtain a dq axis stator current difference delta i under the action of a k moment voltage vector V (k)_dI V (k) and Δ i_q|V(k)；

And 4, step 4: will be Δ i_dI V (k) and Δ i_qI V (k) is sent to a current difference lookup table updating module to calculate the stator current difference corresponding to the voltage vector in the reverse direction of V (k), and the current difference lookup table is updated;

and 5: detecting whether the zero vector is applied in a set sampling period, and if not, forcibly outputting the zero vector as a basic voltage vector for driving the inverter;

step 6: when the zero vector is applied in a set sampling period, a basic voltage vector meeting the minimum cost function is obtained through rolling optimization of the cost function and serves as a voltage vector for controlling the inverter.

Further, in the step 1, a given rotating speed N is set_r ^refAnd the actual speed N measured by the encoder_rDeviation e between_nA PI controller for inputting the rotation speed loop, and calculating a reference value i of the q-axis current according to formula (1)_q ^ref(k+1)；

Wherein k is_pAnd k_iRespectively, proportional gain and integral gain of the rotating speed PI controller, and s is a complex variable.

Further, in the step 2, Clark transformation is shown as formula (2); the Park transformation is shown in formula (3);

wherein i_α(k-1)、i_β(k-1) stator current of α β axis at time (k-1), i_α(k) And i_β(k) Stator current of the α β axis at time k.

Further, in the step 3, the dq-axis current difference calculation module obtains the Δ i according to a formula (4)_dI V (k) and Δ i_q|V(k)；

Further, in step 4, the method for updating the current difference lookup table includes:

step A: judging whether the voltage vector V (k) at the moment k is a zero vector, if so, acquiring the zero vector V according to a formula (5)₀And zero vector V₇Applied dq-axis stator current difference Δ i_x|V₀、Δi_x|V₇X belongs to { d, q }, and updating the current difference lookup table; if V (k) is not a zero vector, performing step B;

and B: will vector V_j(j ∈ {0,1,2,3,4,5,6,7}) under the action of dq axis stator current difference i_x|V_jDivided into zero vector current difference Δ i_x|V₀And a non-zero vector_x|V_jTwo parts, as shown in equation (6), where i_x|V₀＝i_x|V ₇0; the non-zero base voltage vector of the two-level inverter is then divided into 3 sets CVS (1) { V }₁,V₆}，CVS(2)＝{V₂,V₅}，CVS(3)＝{V₃,V₄And judging the voltage vector V corresponding to V (k)_jThen calculated with V according to equation (7)_jCurrent difference Δ i for another voltage vector in the same set_x|V_mWhere m + j is 7, Δ i_x|V₀Is the value stored in the current difference look-up table; finally, will Δ i_x|V_jAnd Δ i_x|V_mSending the current difference lookup table to replace the original value, and finishing updating the current difference lookup table;

where x ∈ { d, q }.

Further, in step 5, the method of forcibly outputting the zero vector as the basic voltage vector for driving the inverter includes: first, the minimum number n of update cycles of the zero vector is set_{0_period}Then, it is judged at n_{0_period}Whether a zero-crossing vector is output in each sampling period is used as a voltage vector for controlling the inverter, and if so, the zero-crossing vector is not output forcibly; otherwise, directly outputting the zero vector as a voltage vector for controlling the inverter.

Further, in step 6, the method for obtaining the basic voltage vector meeting the minimum cost function through the rolling optimization of the cost function as the voltage vector for controlling the inverter includes: root of herbaceous plantCalculating the predicted value i of the current of the dq axis in different switching states at the moment of (k +1) according to the formula (8)_d(k+1)|V_jAnd i_q(k+1)|V_j(ii) a The output g of the cost function is then calculated by equation (9)_j(j ∈ {0,1,2,3,4,5,6,7 }); finally, obtaining the minimum value function output g through the formula (10)_minThen V is_minNamely, the basic voltage vector acted on the inverter;

g_min＝min{g₀,g₁,g₂,g₃,g₄,g₅,g₆,g₇} (10)

wherein i_x|V_jIs a vector V_j(j is set to {0,1,2,3,4,5,6,7}) and x is set to { d, q }.

Has the advantages that: the embedded permanent magnet synchronous motor based on power supply of the two-level inverter replaces a complex prediction model based on motor parameters by establishing a current prediction equation based on current difference and current accumulation operation, effectively improves the parameter robustness of the system and reduces the calculation burden of the system. Secondly, in order to ensure the quick update of the current difference, a multi-current-difference updating method is provided, which can realize the simultaneous update of two different current differences in one control period.

Drawings

FIG. 1 is a control schematic provided by the present invention;

FIG. 2 is a control flow diagram provided by the present invention;

FIG. 3 is a diagram of a multi-current-difference updated PMSM model-free predicted current control steady-state simulation;

FIG. 4 is a simulation diagram of q-axis current tracking performance of multi-current-difference-updated PMSM model-free predictive current control.

Detailed Description

The present invention will be described in further detail below by way of examples with reference to the accompanying drawings, which are illustrative of the present invention and are not to be construed as limiting the present invention.

A schematic diagram of a multi-current-difference-updated model-free predicted current control method of a permanent magnet synchronous motor is shown in figure 1, and the method comprises a rotating speed loop PI controller module 1, a value function module 2, a two-level inverter module 3, a predicted current model 4, a current difference lookup table 5, a dq-axis current difference calculation module 6, a coordinate transformation module 7, a PMSM module 8 and an encoder module 9.

As shown in fig. 2, the method comprises the following steps:

step 1: controlling a reference value i of q-axis current of a PI controller according to a rotating speed loop_q ^ref(k+1)：

Will give a rotation speed N_r ^refAnd the actual speed N measured by the encoder_rDeviation e between_nA PI controller for inputting the rotation speed loop, and calculating a reference value i of the q-axis current at the time of (k +1) according to a formula (1)_q ^ref(k +1) and giving a reference value i of the d-axis current_d ^ref(k+1)＝0；

Wherein k is_pAnd k_iRespectively, proportional gain and integral gain of the rotating speed PI controller, and s is a complex variable.

Step 2: acquiring three-phase stator current i at (k-1) time and k time through a current sensor_s(k-1) and i_s(k) S ═ a, b, c; then, obtaining stator currents of the (k-1) moment and the k moment alpha beta axis through Clark transformation shown in a formula (2); and then, acquiring the rotor electrical angle theta at the moment k through an encoder, and calculating the dq axis stator current i at the moment (k-1) and the moment k through Park change shown in a formula (3)_d(k-1)、i_q(k-1) and i_d(k)、i_q(k)；

Wherein i_α(k-1)、i_β(k-1) stator current of α β axis at time (k-1), i_α(k) And i_β(k) Stator current of the α β axis at time k.

And step 3: will i_d(k-1)、i_q(k-1)、i_d(k) And i_q(k) Sending the current difference to a dq axis current difference calculation module, and obtaining a dq axis stator current difference delta i under the action of a voltage vector V (k) at the k moment according to a formula (4)_dI V (k) and Δ i_q|V(k)；

And 4, step 4: will be Δ i_dI V (k) and Δ i_qI V (k) is sent to a current difference lookup table updating module to calculate the stator current difference corresponding to the voltage vector in the reverse direction of V (k), and the current difference lookup table is updated; the method specifically comprises the following steps:

step A: judging whether the voltage vector V (k) at the moment k is a zero vector V₀Or zero vector V₇If so, V is obtained according to equation (5)₀And V₇Applied dq-axis stator current difference Δ i_x|V₀And Δ i_x|V₇X belongs to { d, q }, and updating the current difference lookup table; if V (k) is not a zero vector, performing step B;

and B: will vector V_j(j ∈ {0,1,2,3,4,5,6,7}) under the action of dq axis stator current difference i_x|V_jDivided into zero vector current difference Δ i_x|V₀And a non-zero vector_x|V_jTwo parts, as shown in equation (6), where i_x|V₀＝i_x|V ₇0; the non-zero base voltage vector of the two-level inverter is then divided into 3 sets CVS (1) { V }₁,V₆}，CVS(2)＝{V₂,V₅}，CVS(3)＝{V₃,V₄And judging the voltage vector V corresponding to V (k)_jThen calculated with V according to equation (7)_jCurrent difference Δ i for another voltage vector in the same set_x|V_mWhere m + j is 7, Δ i_x|V₀Is the value stored in the current difference look-up table; finally, will Δ i_x|V_jAnd Δ i_x|V_mSending the current difference lookup table to replace the original delta i_x|V_jAnd Δ i_x|V_mAnd finishing the updating of the current difference lookup table.

Where x ∈ { d, q }.

And 5: and detecting whether the zero vector is applied in a set sampling period, and if not, forcibly outputting the zero vector as a basic voltage vector for driving the inverter. The method for forcibly outputting the zero vector as the basic voltage vector for driving the inverter comprises the following steps: first, the minimum number n of update cycles of the zero vector is set_{0_period}The present embodiment is set to 100 sampling periods and then judged at n_{0_period}Whether a zero-crossing vector is output in each sampling period is used as a voltage vector for controlling the inverter, and if so, the zero-crossing vector is not output forcibly; otherwise, directly outputting the zero vector as a voltage vector for controlling the inverter.

Step 6: when the zero vector is applied in the set sampling period, the basic voltage vector meeting the minimum cost function is obtained through the rolling optimization of the cost function to be used as a controlThe voltage vector of the inverter is modulated. The method for acquiring the basic voltage vector meeting the minimum cost function through the rolling optimization of the cost function as the voltage vector for controlling the inverter comprises the following steps: firstly, whether a zero vector is output forcibly is judged through the step 5, if not, a predicted value i of the current of the dq axis in different switching states at the moment of (k +1) is calculated according to a formula (8)_d(k+1)|V_jAnd i_q(k+1)|V_j(ii) a The output g of the cost function is then calculated by equation (9)_j(j ∈ {0,1,2,3,4,5,6,7 }); finally, obtaining the minimum value function output g through the formula (10)_minThen V is_minI.e. the basic voltage vector acting on the inverter.

g_min＝min{g₀,g₁,g₂,g₃,g₄,g₅,g₆,g₇} (10)

Wherein i_x|V_jIs a vector V_j(j is set to {0,1,2,3,4,5,6,7}) and x is set to { d, q }.

The simulation results of the method of the present invention are shown in fig. 3 and 4. Fig. 3(a) and (b) show the average current tracking error (MCTE) with 50% increase in dq inductance, 3(c) shows MCTE at nominal parameters, and 3(d) shows MCTE for MFPCC, which shows that since the MFPCC algorithm does not require motor parameters to participate in the operation, it can obtain substantially the same current tracking performance as 3 (c); fig. 4(a) and (b) show the steady-state performance of the MPCC and mfccc algorithms, and it can be seen that the mfccc algorithm can obtain good steady-state performance substantially the same as the MPCC algorithm under ideal parameters.

The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (7)

1. A permanent magnet synchronous motor model-free predicted current control method with multiple current difference updating is characterized by comprising the following steps:

step 1: will give a rotation speed N_r ^refAnd the actual speed N measured by the encoder_rThe deviation between the two is obtained by a PI controller of a rotating speed loop to obtain a q-axis current reference value i at the moment of (k +1)_q ^ref(k +1) and giving a reference value i of the d-axis current_d ^ref(k+1)＝0；

Step 2: acquiring three-phase stator current i at (k-1) time and k time through a current sensor_s(k-1) and i_s(k) S ═ a, b, c; then obtaining stator currents of the (k-1) moment and the k moment alpha beta axis through Clark transformation; and then, acquiring the rotor electrical angle theta at the time k through an encoder, and calculating the dq-axis stator current i at the time (k-1) and the time k through Park change_d(k-1)、i_q(k-1) and i_d(k)、i_q(k)；

And step 3: will i_d(k-1)、i_q(k-1)、i_d(k) And i_q(k) Sending the current difference to a dq axis current difference calculation module to obtain a dq axis stator current difference delta i under the action of a k moment voltage vector V (k)_dI V (k) and Δ i_q|V(k)；

And 4, step 4: will be Δ i_dI V (k) and Δ i_qI V (k) is sent to a current difference lookup table updating module to calculate the stator current difference corresponding to the voltage vector in the reverse direction of V (k), and the current difference lookup table is updated;

and 5: detecting whether the zero vector is applied in a set sampling period, and if not, forcibly outputting the zero vector as a basic voltage vector for driving the inverter;

step 6: when the zero vector is applied in a set sampling period, a basic voltage vector meeting the minimum cost function is obtained through rolling optimization of the cost function and serves as a voltage vector for controlling the inverter.

2. The method as claimed in claim 1, wherein in step 1, a given rotating speed N is set_r ^refAnd the actual speed N measured by the encoder_rDeviation e between_nA PI controller for inputting the rotation speed loop, and calculating a reference value i of the q-axis current according to formula (1)_q ^ref(k+1)；

Wherein k is_pAnd k_iRespectively, proportional gain and integral gain of the rotating speed PI controller, and s is a complex variable.

3. The model-free predictive current control method for the multiple current difference updating permanent magnet synchronous motor according to claim 1, wherein in the step 2, Clark transformation is shown as a formula (2); the Park transformation is shown in formula (3);

wherein i_α(k-1)、i_β(k-1) stator current of α β axis at time (k-1), i_α(k) And i_β(k) Stator current of the α β axis at time k.

4. The method as claimed in claim 1, wherein in step 3, the dq-axis current difference calculation module obtains the Δ i according to formula (4)_dI V (k) and Δ i_q|V(k)；

5. The method for controlling the model-free predicted current of the multiple current difference updating permanent magnet synchronous motor according to claim 1, wherein in the step 4, the method for updating the current difference lookup table comprises the following steps:

step A: judging whether the voltage vector V (k) at the moment k is a zero vector, if so, acquiring the zero vector V according to a formula (5)₀And zero vector V₇Applied dq-axis stator current difference Δ i_x|V₀、Δi_x|V₇X belongs to { d, q }, and updating the current difference lookup table; if V (k) is not a zero vector, performing step B;

and B: will vector V_j(j ∈ {0,1,2,3,4,5,6,7}) under the action of dq axis stator current difference i_x|V_jDivided into zero vector current difference Δ i_x|V₀And a non-zero vector_x|V_jTwo parts, as shown in equation (6), where i_x|V₀＝i_x|V₇0; the non-zero base voltage vector of the two-level inverter is then divided into 3 sets CVS (1) { V }₁,V₆}，CVS(2)＝{V₂,V₅}，CVS(3)＝{V₃,V₄And judging the voltage vector V corresponding to V (k)_jThen calculated with V according to equation (7)_jCurrent difference Δ i for another voltage vector in the same set_x|V_mWhere m + j is 7, Δ i_x|V₀Is the value stored in the current difference look-up table; finally, will Δ i_x|V_jAnd Δ i_x|V_mSending the current difference lookup table to replace the original value, and finishing updating the current difference lookup table;

where x ∈ { d, q }.

6. The method for controlling the multiple current difference updating permanent magnet synchronous motor model-free predicted current according to claim 1, wherein in the step 5, a method for forcibly outputting a zero vector as a basic voltage vector for driving an inverter comprises the following steps: first, the minimum number n of update cycles of the zero vector is set_{0_period}Then, it is judged at n_{0_period}Whether a zero-crossing vector is output in each sampling period is used as a voltage vector for controlling the inverter, and if so, the zero-crossing vector is not output forcibly; otherwise, directly outputting the zero vector as a voltage vector for controlling the inverter.

7. The method for controlling the current of the multiple current difference updating permanent magnet synchronous motor without the model prediction as recited in claim 1, wherein in the step 6, the method for obtaining the basic voltage vector meeting the minimum cost function through the rolling optimization of the cost function as the voltage vector for controlling the inverter comprises the following steps: calculating a predicted value i of the current of the dq axis in different switching states at the moment (k +1) according to the formula (8)_d(k+1)|V_jAnd i_q(k+1)|V_j(ii) a The output g of the cost function is then calculated by equation (9)_j(j ∈ {0,1,2,3,4,5,6,7 }); finally, obtaining the minimum value function output g through the formula (10)_minThen V is_minNamely, the basic voltage vector acted on the inverter;

g_min＝min{g₀,g₁,g₂,g₃,g₄,g₅,g₆,g₇} (10)

wherein i_x|V_jIs a vector V_j(j is set to {0,1,2,3,4,5,6,7}) and x is set to { d, q }.

Images